April
1998
given by nasal cannula. Full monitoring, including pulse oximetry, was employed. Results: Sedation was induced in 47/49 children; 2 remained awake. Despite "standard" doses of medications, sedation levels were observed to vary greatly from patient to patient and within the same patient. 22/49 (44%) entered deep sedation at some point during the procedure. 30•47 children (63%) were observed to experience mild or no discomfort; 15/47 children (31%) had moderate or severe discomfort. COL was more painful than EGD (p < 0.05). Amnesia for pain was complete; no patients recalled any pain. Patient and parent satisfaction were 80% and 74% respectively. The observing physician was satisfied with sedation in 27147 children (57%), whereas 20•47 procedures (43%) received lower ratings due to patient discomfort or agitation. Physician satisfaction was better for EGD than COL (p < 0.05); discomfort during COL occasionally interfered with completion of the procedure. Two patients had brief episodes of mild hypotension, one had prolonged recovery from sedation > 3hrs, there were no 02 desaturations. Neither patients' age nor total medication dosage correlated with level of pain control, amnesia, or satisfaction. Conclusions: In this series CS was safely employed in children with good levels of satisfaction among patients and parents. Amnesia for the procedure was excellent. However, levels of observed pain and agitation during EGD and especially COL, indicate lessthan-adequate sedation and pain control. These problems and the high frequency with which undesirable deep sedation occurred raise fundamental questions concerning the appropriate use of this form of CS in children. • G0033 ANTIBIOTIC PROPHYLAXIS FOR SPONTANEOUS BACTERIAL PERITONITIS IN CIRRHOSIS: A STRATEGY OF RISK STRATIFICATION IMPROVES COST-EFFECTIVENESS. A Das. Division of Gastroenterology, University Hospitals of Cleveland, Cleveland, Ohio. Introduction: Selective Intestinal decontamination with different antibiotics have been shown to be effective in preventing most episodes of SBP and in general, thought to be a cost-effective strategy. Since the risk factors for developing an episode of SBP are well identified, it is important to know if restricting antibiotic prophylaxis to a subgroup of patients with cirrhosis and ascites who are at high risk for SBP will improve cost-effectiveness. Methods: Two different strategies of antibiotic prophylaxis were compared with a 'no prophylaxis' strategy in patients with cirrhosis and ascites who are at risk for developing SBP using a decision analysis model. In strategy I, antibiotic prophylaxis was administered in all patients with cirrhosis and ascites and in strategy II, patients were stratified in to a low risk and a high risk group on the basis of a serum bilirubin and ascitic fluid protein levels; only patients in the high risk group (as identified by serum bilirubin > 2.5 mg/dl and ascitic fluid protein < 1 Gm/dl) received antibiotic prophylaxis. The cost per patient treated over one year was the outcome measure compared in the different strategies. Clinical input probabilities and ranges used were obtained by searching the MEDLINE database for English language articles. Cost estimates were obtained from a university hospital setting, cost analysis was done with a societal perspective and only direct costs were taken in to account. Sensitivity analyses were performed to evaluate the effect of variations in the key clinical probabilities and cost estimates ranging from a best case to worst case scenario on the outcome measure. Results: The estimated cost per patient treated in strategy I, strategy II and the strategy of 'no prophylaxis' (strategy III) were $1311, $1123 and $3509 respectively. Over a broad range of clinical and cost variable, the strategy of risk stratification and restriction of antibiotic prophylaxis to the subgroup of patients with cirrhosis and ascites who are at high risk for SBP was the most favored strategy. However, when the cost of prophylaxis is low or the probability of a primary episode of SBP is at the lower end of the range reported in literature, administering antibiotic prophylaxis to all patients with cirrhosis and ascites is the least costly strategy. Conclusion: This cost analysis indicates that antibiotic prophylaxis is generally cost-effective in the prevention of SBP in patients with cirrhosis and ascites. The strategy of restricting antibiotic prophylaxis to the subgroup of patients who, by simple laboratory parameters, are identified to be at high risk of SBP significantly improves cost-effectiveness. • G0034 SHOULD LIVER BIOPSY BE DONE IN ASYMPTOMATIC PATIENTS WITH CHRONICALLY ELEVATED TRANSAMINASES: A COSTUTILITY ANALYSIS. A Das, AB Post. Division of Gastroenterology, University Hospitals of Cleveland, Cleveland, Ohio Introduction: A percutaneous liver biopsy is recommended in the evaluation of asymptomatic patients with chronic elevation of serum transaminases and negative diagnostic work up. Such a recommendation is based on older studies which do not specifically take in to account the prevalence of chronic hepatitis C infection. Since patients with Hepatitis C related chronic liver disease can now be diagnosed with noninvasive tests, the diagnostic and therapeutic utility of an invasive procedure like liver biopsy in asymptomatic subjects warrants rigorous scrutiny. Methods & Materials: A decision analysis model was constructed by using DATA 3.0. software (Treeage Softwares, Inc.) to study the cost-utility of liver biopsy in a hypothetical cohort of asymptomatic adult patients with chronic ( > 6 months) mild to moderate elevation of transaminases. Patients with comorbid illness, alcoholism and high risk behavior for HIV infection were excluded. All patients had negative viral and autoimmune serological markers, as well as studies for metabolic disorders. Two management
Clinical Practice A9 strategies were compared in this model; in strategy I, a liver biopsy was done in all patients at the beginning and appropriate therapeutic decisions were made based on the histopathologic finding. In strategy II, after an initial evaluation patients were followed expectantly for a period of at least ten years with annual physical examinations and liver function tests; at any point if they became symptomatic they were further investigated as needed but were no longer included in the decision model. Cost of management and quality adjusted life years (QALY)s were the two outcome measures considered. For baseline analysis, it was assumed on the basis of available literature that the proportion of patients in the model cohort with fatty, alcoholic, necroinflammatory and other miscellaneous disorders were 0.25, 0.25, 0.3 and 0.2 respectively. Clinical probabilities were obtained from the available literature. Cost estimates were from a university hospital setting, only direct costs (discounted at 5%) were calculated and the analysis was done with a societal perspective. Results: Results of the baseline analysis for a fifty year old person were as follows: Strategy Strategy 1 Strategy II
Cost of management ($) $1625 $1325
QALY 24 25
C/E (cost per QALY) $67 $53
In sensitivity analyses done with key clinical probabilities and cost estimates varying from a best case to a worst case scenario, the strategy of expectant follow up was dominant. The proportion of patients in the clinical follow up group who would eventually be diagnosed and treated strongly influenced the results of the analysis. Conclusion: Expectant clinical follow up is the most cost-effective strategy in managing asymptomatic patients with negative viral, metabolic and autoimmune markers and chronically elevated serum transaminases. G0035 PROSPECTIVE RANDOMISED STUDY OF THE USE OF EARLY LAPAROSCOPY IN THE MANAGEMENT OF ACUTE NONSPECIFIC ABDOMINAL PAIN (NSAP). B. Decadt. L. Sussman, M. Lewis, L. Cohen, A. Patel, C. Rogers & M. Rhodes. Norfolk & Norwich Hospital, Norwich, England.
Early laparoscopy may improve management of NSAP. Between November 1995 and November 1997, 99 patients with NSAP were randomised into early laparoscopy within 18 hours (46 patients) versus a traditional "wait and see" policy (53 patients). We aimed to compare hospital stay, morbidity, mortality, diagnosis, readmission rate after 6 months follow-up and gastrointestinal and psychological well-being (scored on admission and after 6 weeks). Results: Early laparoscopy wait & see age 28 (range 16-84) 28 (range 16-60) hospital stay 2.5 days 2.5 days morbidity 10 (21.7%) 12 (22.6%) mortality 1 1 diagnosis* 41/46 20/51 re-admission 3/28 patients 11/36 patients within 6/12 well-being 133/177~149/177 132/177~144/177 score (p = 0.004) (p = 0.0132) (*2 pts were excluded after breaching the protocol)
ns ns ns ns p < 0.001 p < 0.05 ns
Early laparoscopy provided a lower re-admission rate and a higher diagnostic accuracy in patients with NSAP. There was no significant difference in hospital stay, patient well-being, morbidity or mortality. G0036
STAGING OF PANCREATIC AND HEPATOBILIARY MALIGNANCIES BY LAPAROSCOPY AND LAPAROSCOPIC ULTRASOUND. B. Decadt L. Cohen, L. Sussman, A. Patel, M. Watson, M. Lewis & M. Rhodes. Norfolk & Norwich Hospital, Norwich, England. Laparoscopy and laparoscopic ultrasound may optimize patient selection for cancer resection with curative intent. From August 1995 to November 1997, 68 patients (30F: 38M) aged 30-88 (median age 63) underwent laparoscopic staging for mainly upper GI malignancy. Laparoscopic ultrasound was used in 47 patients. No evidence of cancer was found in 11 patients and 1 laparoscopy failed due to adhesions. 32 patients were found not suitable for curative surgery because of peritoneal involvement (11 pts.), hepatic involvement (14 pts), portal vein or coeliac plexus encasement (6 pts) or advanced local pancreatic disease (1 pt.). The remaining 24 patients underwent laparotomy with curative intent. This failed in 7 patients due to a peroperative cardiac event (1 pt.), an unexpected T-stage (3 pts), an unexpected N-stage (1 pt) or bilateral liver involvement (2 pts). Curative resections involved 7 hepatectomies, 5 Whipple's procedures and 5 other procedures. Laparoscopic ultrasound modified therapeutic planning in 18147 (38.3%) pts. Laparoscopy changed staging in 32/68 (47%) of the cases. Laparoscopy identified resectability with a sensitivity of 100% and a specificity of 82%. A laparotomy was avoided in 37/68 (54%) pts. Laparoscopy and laparoscopic ultrasound are valuable tools in the management of upper GI malignancies.